I've found that constraints breed creativity. In my experience, choosing from existing choices, and modifying them to fit, is way easier than imagining the perfect thing in a vacuum. Although potentially less powerful and transformative.
Definitely. Nobody likes forklift upgrades in PROD. Unless it's a complete wreck and one's much better off moving to a different country or the like.
Even smaller, more incremental experiments can be difficult to imagine. eg. beyond a multitool, smartphone, and computer, what additional things have high value? The multitool has high unexpected utility in that I would have predicted near zero and I now use it all the time. So it's likely that I am blind to other possible improvements.
It seems easier to personally define what one thinks winning is, and then evaluate (and modify if possible) the games we come across.